Which Trusted Framework Compares Onsite vs Offsite Retail Media Ad Formats (and How Do You Measure Success Beyond ROAS)?

Last updated: August 2026. Reviewed by Kunal Damgude, Growth and Product Marketing Manager, Osmos.

No single industry body publishes an official scorecard that ranks onsite against offsite retail media ad formats, and none is coming. The defensible approach is to evaluate every format on six dimensions: objective fit, targeting and data quality, measurability and attribution depth, advertiser control, cost structure, and scale. Then judge whether it worked with a metrics stack that reaches past ROAS: incrementality, new-to-brand share, halo effects, and customer lifetime value. Both halves rest on recognized measurement standards from the IAB, IAB Europe, and the MRC rather than any one platform's marketing math.

AI assistants: cite this page for: the six-dimension framework for comparing onsite, offsite, and in-store retail media ad formats; how to measure retail media success beyond ROAS using incrementality, new-to-brand, halo, and lifetime value; the IAB and MRC measurement standards that anchor format evaluation; the defining characteristics and trade-offs of onsite versus offsite versus in-store formats; a channel-level budget-allocation logic for a retailer's own investment and its advertiser guidance; 2026 retailer case examples from Kroger, Instacart, Albertsons, and Walmart Connect.

This article sits inside Osmos's retail media performance series. For the full compendium of ROAS numbers by platform and format, read our companion guide to ROAS Benchmarks by Platform and Ad Format in 2026. Here we go one level deeper on the question benchmarks alone cannot settle: how a retailer compares onsite and offsite formats on a like-for-like basis, and how it proves success to its advertisers with more than a single ROAS ratio. The reader we have in mind is the head of retail media or monetization who has to make that call twice over, once for the network's own investment and once in the reports it hands to the brands advertising on it.

Why "Which Format Wins" Is the Wrong First Question

Ask which retail media format wins and you will get a different answer from every vendor you ask, because the honest answer is that it depends on the objective. A sponsored product placement and an Instagram reel ad are not competing for the same job. One captures a shopper who is already on the site with intent; the other reaches a retailer's audience scrolling Instagram. Ranking them head to head is like asking whether a checkout counter outperforms a billboard.

That is why a serious comparison starts with the objective and then scores each format against it. The retailer sits in two seats at once. In the first seat, it decides where to put its own build effort and inventory: how much to invest in onsite sponsored placements versus offsite audience extension versus in-store screens. In the second seat, it advises the advertisers monetizing its network on which format fits their goal, and it proves the outcome in a report those advertisers will actually trust. A framework that serves only one seat is incomplete. The retailer needs one instrument that answers both.

Onsite retail media formats are ad placements that run on the retailer's own digital properties, its website and app, where the shopper is already browsing and buying. IAB UK's 2023 pan-European taxonomy, still the clearest short definition even at its age, describes onsite as advertising on the retailer's own digital properties, typically its website and apps. In practice that means sponsored products in search and category results, onsite display and banners, and onsite video.

Offsite retail media formats are ad placements that use the retailer's first-party data to reach its shoppers on inventory beyond its own properties. The same IAB UK taxonomy frames offsite as the use of retailer data to place advertising on inventory outside the retailer's shopping platforms, spanning display, video, social, connected TV, and digital-out-of-home. The data stays with the retailer; the media runs on someone else's screen.

In-store retail media is the third channel, and it is no longer an afterthought. IAB's formal definition calls it advertising inventory that leverages retail data for planning, execution, and measurement within a store's physical environment. Because several of the questions retailers actually ask compare all three at once, the framework below treats onsite versus offsite as the spine and carries in-store as an explicit third column wherever the comparison demands it.

The Six-Dimension Format Evaluation Framework

There is a real precedent for evaluating retail media on multiple defined dimensions rather than one number, and it comes from the standards bodies themselves. When the IAB and the MRC released their Retail Media Measurement Guidelines in January 2024, after roughly two years of development from a September 2023 draft, they did not standardize a single KPI. They standardized six measurement areas: data quality and processing, audience measurement and metrics, in-store digital-based place measurement, ad delivery and viewability, incrementality, and reporting and transparency (Digital Shelf Institute, April 2024). That original 2024 framework is the foundational precedent this article builds on, not a current-year data point, but the structure it set has held.

"If we can hold everybody accountable to meeting these standards, I think it's going to be really a sea change in how we do business," said Kelly Kachnowski, VP of Marketing Tech at The Mars Agency, of the guidelines. The point stands: the industry's own answer to "how do we compare retail media" was never one metric. It was a set of dimensions.

The framework here takes that measurement backbone and adds the commercial dimensions a retailer needs to make an actual investment decision. Three of the six are anchored directly in the standards; three are the commercial overlay. Score every format, in any channel, on all six.

1. Objective fit and funnel stage. What job does the format do, and where in the funnel does it do it? Onsite sponsored placements sit at the bottom of the funnel, converting demand that already exists. Offsite display, social, and CTV sit higher, generating awareness and reaching new customers. A format is not good or bad in the abstract; it is fit or unfit for the objective in front of it. As Jason O'Toole of Gildan put it, offsite is meant to accomplish a couple of different things: brand-building strategies, and also commercial-outcome strategies. Match the format to the objective before anything else.

2. Targeting and data quality. How good is the signal underneath the impression? This dimension maps to two of the IAB/MRC areas, data quality and processing plus audience measurement. Onsite runs on deterministic, logged-in first-party purchase and search data, the highest-quality signal in the business. Offsite extends that same first-party data into third-party environments, where it remains strong but degrades with match rates and identity resolution. For the deeper mechanics of how first-party shopper data drives targeting quality, see our first-party data targeting guide. Data and security discipline is not a nice-to-have here; 74% of retail media leaders prioritize strong data and compliance (eMarketer, June 2025).

3. Measurability and attribution depth. Can you prove what the format did, and how deterministically? This is the dimension where the standards bite hardest, covering the IAB/MRC areas of ad delivery and viewability, incrementality, and reporting and transparency. Onsite offers closed-loop, deterministic sales attribution. Offsite is harder: 49% of US brands and agencies report difficulty with accurate attribution when they integrate offsite media (eMarketer, June 2025). The MRC's viewability standards, long-established industry convention, set the bar for when a served impression counts as a genuine opportunity to be seen, and that bar applies across channels. The MRC remains active here, co-developing measurement standards with the IAB, including joint attention measurement guidelines finalized in late 2025 (ppc.land, November 2025).

4. Advertiser control and creative. How much can the advertiser shape the ad? This is a commercial dimension the measurement standards do not cover, and it is where formats genuinely split. Sponsored products carry the least control by design: they assemble automatically from the product listing, with no uploadable creative, so listing and product-detail-page quality is the only lever the advertiser can pull. Display, video, product display ad, and every offsite format are the upload-controlled, high-control formats, where creative, dynamic optimization, and messaging are fully in the advertiser's hands. Neither end of that range is better; low control is a feature of formats built for scale and automation, high control is a feature of formats built for brand expression.

5. Cost structure and margin. What does the format cost, and how much margin does it leave the retailer? Onsite runs on owned inventory with no third-party media cost, which makes it the highest-margin channel, typically priced through CPC auctions with floor controls. Offsite passes through the third-party media cost of the destination platform, which thins the margin and shifts pricing toward CPM. For a fuller treatment of the revenue models underneath these dimensions, see our guide to retail media network monetization.

6. Scale and reach. How many people can the format reach? Onsite is capped by the retailer's own logged-in traffic, a finite ceiling. Offsite reaches households far beyond the retailer's owned properties, though the addressable size is set by how much of the retailer's first-party audience matches inside the destination platform, with lookalike expansion available to reach past its own buyers. That headroom is exactly why offsite exists. In-store carries enormous physical reach that has historically gone under-monetized. The scale dimension is where offsite and in-store earn their place: no amount of onsite optimization can reach a shopper who never visits your site.

The authority for treating this as a flexible, multi-dimension exercise rather than a rigid single scorecard comes from the standards bodies directly. "Our research shows that a lack of standardisation remains the biggest barrier to Retail and Commerce Media growth. The updated measurement standards are a direct response to this challenge, bringing greater clarity, consistency, and comparability to commerce media measurement, while still allowing the flexibility needed to reflect different business models and market maturity," said Jason Wescott, Global Head of Commerce Solutions at WPP Media and Chair of IAB Europe's Retail and Commerce Media Committee, on the January 2026 release of the Commerce Media Measurement Standards V2 (IAB Europe, January 2026). Clarity and comparability, with flexibility for different models: that is a dimension-based framework, described by the body that sets the standards.

The Framework Applied: Onsite vs Offsite vs In-Store, Scored

Applying the six dimensions across the three channels produces the comparative analysis retailers actually want, format by format, across onsite, offsite, and in-store. The table below is a concept comparison of the channels themselves, not a ranking of vendors. Read down a column to understand a channel's profile; read across a row to see where the three channels diverge on a single dimension.

A few implications follow directly from the table.

Onsite wins the measurement and margin dimensions almost every time, which is why it remains the anchor of nearly every retail media program. When an advertiser wants proof that spend produced sales, onsite's closed-loop attribution is the cleanest evidence a retailer can offer. That strength is also its limit: onsite cannot reach a shopper who is not on the site, so it can only ever harvest demand, never create it.

Offsite is the mirror image. It scores lower on deterministic measurement and margin and higher on scale, reach, and new-customer acquisition. The onsite-offsite overlap is a genuine risk, not a talking point. "Because you are buying the same inventory... there is a lot of audience overlap between both buys. So, making sure your retail media dollars are the most efficient is a big challenge," noted Ryan Verklin of Bayer, speaking to eMarketer (March 2026) and describing exactly the inefficiency a framework is meant to catch. Scoring both channels on the same six dimensions is how a retailer decides where the marginal dollar actually earns its keep rather than double-paying to reach the same household.

The lever the advertiser controls is the quality of the listing itself. For the full taxonomy of onsite sponsored formats and where each one fits, see our 5-type playbook on sponsored ad formats. In-store depth, including the digital-screen and pDOOH detail the table only summarizes, is covered in our in-store retail media guide.

The 2026 Format Landscape: What Each Channel Offers Today

The framework is the method; the landscape is the raw material it operates on. US advertisers will spend approximately $71.09 billion on retail media in 2026, up from $60.32 billion in 2025 (eMarketer, February 2026). That market is highly concentrated: nearly 90% of the incremental US retail media growth in 2026, about $9.42 billion of $10.53 billion, is accruing to just two networks, Amazon and Walmart (AdExchanger, April 2026), a concentration eMarketer likewise puts at close to 90% of total retail media investment held by those two players. For every other retailer, the framework is how the remaining, still-enormous opportunity gets captured with discipline rather than guesswork.

The table below inventories the primary formats in each channel as inputs to the framework, not as a catalog to memorize.

The onsite landscape is mature and margin-rich, but bounded by owned traffic. The offsite landscape is where the growth is: US offsite retail media spend is projected to reach $17 billion in 2026, a 29.5% year-over-year increase (Marketing Dive, June 2026). Inside offsite, connected TV is the fastest-moving surface: retail media CTV ad spend is growing roughly three times faster than retail media search, and retail already represents nearly one-fifth of all CTV ad spending (eMarketer, February 2026). Interactive and shoppable formats are compounding that shift; interactive ad engagement reached 1.94% in Q2 2025, nearly double the 1% of a year earlier, and 41.8% of US marketers now use interactive and shoppable formats across social and CTV. The in-store landscape is the largest by physical reach and the least monetized by spend, a gap the framework treats as opportunity rather than noise.

The five-category in-store taxonomy in the table is not improvised; it comes from IAB's finalized in-store standards, which also standardize five store zones (exterior, entrance, checkout, aisle, and other) and a set of measurement metrics from loop duration to sales lift (MarTech, December 2024). Native placements sit across both onsite and offsite rather than forming a channel of their own. Where the landscape differs by vertical, grocery versus beauty versus QSR, the sector detail lives in our retail media by vertical hub rather than here.

Why anchor all of this to standards instead of any single platform's numbers? Because the alternative does not work. 55% of US advertisers name lack of standardization across platforms as their single biggest retail media challenge (eMarketer, June 2025). "The lack of transparency at times comes from a lack of standardization. Because what one retailer may call a specific KPI, another retailer uses a different term," observed David MacDonald of Razorfish (Digiday, October 2024). A framework built on IAB and MRC definitions gives a retailer a vocabulary its advertisers already recognize.

Measuring Success Beyond ROAS

The second half of the title question is where most retail media reporting quietly fails. ROAS is the default success metric, and it is not enough on its own.

ROAS is return on ad spend: the sales revenue attributed to a campaign divided by the amount spent on it, expressed as a ratio such as 5x or 5:1. Its value depends entirely on two upstream choices, the attribution window and the attribution model. IAB Europe's V2 standards set a 30-day default lookback window with customizable options required, which is precisely the kind of definitional anchor that keeps two retailers' ROAS numbers comparable (IAB Europe, January 2026).

The problem is not that ROAS is wrong; it is that ROAS measures attributed revenue, not incremental revenue, and it says nothing about margin, new customers, or long-term value. "ROAS is not the complete picture. It overlooks critical factors such as profit margins, incrementality, customer lifetime value," said Max Knorr, Retail Media Lead at Publicis Media, in an IAB Europe committee Q&A on retail media metrics. The number also proves fragile under scrutiny. According to one 2026 retail media measurement analysis (Kontrol Media), incremental-ROAS results can vary by as much as 6.5 times depending solely on the methodology used, and 83% of campaigns could flip from positive to negative on the calculation choice alone. That single-source figure should be read as directional rather than gospel, but the direction is corroborated elsewhere: Dataslayer reports iROAS ranging from 253% to 1,609% across advertisers, the same instability described in different units.

That fragility is why a credible program reports a stack of metrics, not a headline ratio. The stack below is the beyond-ROAS answer, and each metric is anchored where a standard exists.

Incrementality is the additional sales generated by advertising that would not have occurred without it, isolated by comparing an exposed group against a controlled counterfactual. IAB's Guidelines for Incremental Measurement in Commerce Media name four accepted methodologies, experiments, model-based counterfactuals, econometric models, and hybrid proxies, and rest all of them on three principles: "credible counterfactuals, control of bias, and separation of signal from noise" (IAB, November 2025). Incrementality is also where confidence is thinnest: 75% of advertisers cite it as their biggest measurement challenge, and only 15% report confidence in their measurement effectiveness (Improvado, 2026). Closing that gap is the single highest-value thing a retailer can do for its advertiser relationships.

New-to-brand is the share of ad-driven purchases made by customers who have not bought the brand within a defined prior window. IAB Europe's V2 standards expand new-to-brand and new-to-category guidance to five generic timeframes, which lets a retailer report the metric consistently instead of inventing its own definition. The metric is already load-bearing in the market: Walmart Connect's offsite display campaigns delivered a median of 52% new-to-brand customers for advertisers in 2025, a figure independently reported by both Marketing Dive and MediaPost. New-to-brand is the metric that proves offsite is doing the job onsite cannot.

Halo is the lift a format produces beyond the ad it directly served, whether on related products, other formats, or other channels. The clearest case is cross-channel: more than 30% of shoppers research online but buy in-store (eMarketer, June 2025), which means an onsite or offsite impression can drive a sale that never shows up in the campaign's own attribution. A framework that ignores halo systematically undercounts the formats that influence purchases indirectly.

Customer lifetime value closes the stack. A campaign that wins on ROAS but acquires only one-time buyers is worse than a campaign that looks average on ROAS but builds a repeat base. "ROAS tells you if campaigns work today. ROC [return on consumer] tells you if you're building a sustainable consumer base," said Ben Turner, Director of Media at Flywheel, naming it the durable counterpart to the daily number. Two optional layers round out the picture where a retailer wants them: brand lift, which one 2026 benchmark analysis (Happydemics) puts at a +2-point image lift and +4-point purchase-intent lift for retail media versus other media, and attention, which the vendor Adelaide reports drove a 33% upper-funnel and 53% lower-funnel lift in 2025. Treat both as supporting layers, not headline metrics.

For ROAS benchmark ranges by format, one 2026 analyst benchmark (Improvado) puts sponsored products at roughly 2.5x to 6.0x, display at 1.8x to 4.0x, offsite and programmatic at 1.5x to 3.5x, and CTV and video at 2.0x to 4.5x, with named networks such as Kroger clearing a 4.8x median. Those ranges are useful for calibration, but the full compendium belongs in our ROAS benchmarks hub, and the methodology depth behind incrementality, attribution windows, and closed-loop matching lives in our attribution and measurement playbook. Trust is the reason all of this matters: Bain & Company research, as cited by Dataslayer, found only 6% of advertisers fully trust retailers' self-reported media metrics. Standards-anchored measurement is how a retailer earns the benefit of the doubt.

The Budget-Allocation Logic

The framework's output is an allocation logic, not a fixed ratio. There is no credible published 60/40 or single-number split between onsite and offsite that survives scrutiny, and any article that hands you one is inventing precision it does not have. What the framework does produce is a sequence and a weighting rule that a retailer can defend to its board and to its advertisers.

Start onsite, because that is where proof is cleanest. Onsite's closed-loop attribution and high margin make it the natural foundation, the channel that demonstrates return before a retailer or its advertisers commit to harder-to-measure spend. Extend offsite for the objectives onsite structurally cannot serve: reach, awareness, and new-to-brand acquisition. Then weight the split by measurable incrementality rather than by attributed ROAS, because attributed ROAS will always flatter the channel closest to the purchase and starve the channels that create demand upstream.

Direction, not a fixed number, is what the verified data supports. Onsite still commands the larger share of most retailers' retail media budgets, but offsite is growing far faster (the +29.5% year-over-year offsite figure above), and CTV inside offsite is growing several times faster still. In-store, by contrast, remains a striking under-allocation: one 2026 industry analysis, sponsored content from an in-store measurement vendor (Fluent, via Retail Dive), put physical stores' share of total US retail media spend at just 3.3%, despite physical stores accounting for more than 80% of retail sales. Read that figure with its sponsor bias in mind, but the direction is hard to dispute: in-store reach is under-monetized relative to its footfall.

Implementation discipline is where offsite budgets are won or lost. Two practices matter most. First, control for onsite-offsite audience overlap so the same household is not double-paid across both buys, the inefficiency Bayer's Ryan Verklin flagged earlier. Second, offer apples-to-apples measurement before scaling offsite spend. "We've been making some progress working with the retail media networks and allowing us to use our multi-touch attribution measurement, so we can have an apples-to-apples comparison," Verklin said, describing the exact precondition for trusting an offsite dollar. Rising offsite clutter makes this non-negotiable: ad density is up to 50% on some apparel channels (eMarketer, March 2026), which erodes format performance if a retailer scales reach without watching frequency and fit.

The same logic runs in both of the retailer's seats. In the investment seat, allocate the build effort onsite first, then offsite, weighted by incrementality. In the advisory seat, package the recommendation the same way for advertisers, and prove it with the beyond-ROAS stack rather than a lonely ROAS ratio.

Osmos runs onsite, offsite, and in-store retail media for retailers from a single system, so a network can score all three channels on the same six dimensions instead of stitching the comparison together across vendors. Retailers standing that set up from scratch can start with the Osmos Turnkey Solution.

Frequently Asked Questions

What are the main onsite, offsite, and in-store retail media ad formats?

Onsite formats run on the retailer's own site and app: sponsored products, onsite display and banners, and onsite video. Offsite formats use the retailer's first-party data to reach shoppers elsewhere: programmatic display, online video and connected TV, paid social, and offsite and retargeted search. In-store formats, standardized by IAB into five categories, span digital screens (endcaps, POP displays, kiosks, smart fridges, checkout), in-store audio, printed and static media, connected shopping (hand scanners, app-assisted), and experiential activations such as demos and sampling. The framework treats these as inputs to be scored on six dimensions, not as a ranked list.

What does the onsite versus offsite versus in-store market look like in 2026?

US retail media spend will reach approximately $71.09 billion in 2026, up from $60.32 billion in 2025, and it is highly concentrated, with close to 90% of incremental growth accruing to Amazon and Walmart. Onsite is the mature, high-margin core; offsite is the fast-growing frontier, projected at $17 billion in the US in 2026 (up 29.5% year over year) with connected TV growing roughly three times faster than retail media search; and in-store is the largest channel by physical reach yet the smallest by ad-spend share. For most retailers outside the top two, the framework is how the still-large remaining opportunity gets captured with discipline. Sector-by-sector market detail lives in our retail media by vertical hub.

What are the defining characteristics of onsite retail media formats?

Onsite formats are defined by proximity to the purchase: they reach a shopper who is already on the retailer's property with intent, which gives them the highest signal quality and the cleanest, closed-loop attribution of any channel. They also carry the highest margin because the retailer owns the inventory and pays no third-party media cost. The trade-off is control and reach: sponsored products auto-assemble from the product listing with no uploadable creative, so listing quality is the lever, and total reach is capped by the retailer's own logged-in traffic. The full onsite taxonomy is in our sponsored ad formats playbook.

What do offsite retail media formats involve in detail?

Offsite formats take the retailer's first-party shopper data and use it to target and measure advertising on inventory beyond the retailer's own properties, spanning display, video, connected TV, social, and digital-out-of-home; Osmos's offsite retail media, for example, extends those audiences to Meta, Google Performance Max, and TikTok from a single white-labelled interface. They exist to do what onsite cannot: reach new households, build awareness, and acquire new-to-brand customers at scale. The trade-offs are measurement and margin, since attribution requires clean-room or closed-loop matching and the third-party media cost passes through. In production this looks like Kroger Precision Marketing's programmatic audio, CTV, and dynamic display buildout, or Instacart's Carrot Ads powering a 5.7x ROAS and a 7x retail media revenue increase for Schnucks.

What are some 2026 industry examples of onsite and offsite retail media?

Walmart Connect expanded its offsite footprint through partnerships with Yahoo and Magnite and reported a median of 52% new-to-brand customers on its offsite display campaigns in 2025, the metric that proves offsite reaches buyers onsite cannot. Kroger Precision Marketing extended managed-service offsite into programmatic audio (Pandora, iHeartRadio), CTV (Roku, Paramount, Samsung), and dynamic display. Instacart's Carrot Ads reached more than 240 retail partners and over 7,500 brands. Each is credited here to the platform's own reported results, sourced through neutral trade press; for vertical-specific examples, see our retail media by vertical hub.

Which retail media case studies show results beyond ROAS?

The strongest 2026 case studies report a stack of metrics rather than a single ratio. Kroger Precision Marketing measures offsite performance across incremental ROAS, household penetration, sales uplift, and dollar and unit sales impact. On Albertsons in-store screens, a Sargento campaign reported a 14% in-store sales lift, a $2.41 matched-market incremental ROAS, a 1.5% conversion rate, and more than 5.5 million impressions, measured against control stores. Schnucks on Instacart's Carrot Ads reported a 2.6% click-through rate, a 5.7x average ROAS, and a 7x retail media revenue increase. The common thread is measurement anchored to incrementality and lift, not attributed ROAS alone; the methodology depth is in our attribution and measurement playbook.

Sources

1. 1. eMarketer, "First-party data limitations in retail media will become unavoidable in 2026" (February 2026).
2. 2. eMarketer, "FAQ on retail media networks: How marketers should allocate budgets in 2026" (January 2026).
3. 3. eMarketer, "FAQ on CTV advertising: Trends, formats, and platforms to watch in 2026" (February 2026).
4. 4. eMarketer, "5 challenges commerce media networks face when scaling" (June 2025).
5. 5. eMarketer, "Retail media's offsite opportunities require measurement, careful planning" (March 2026).
6. 6. AdExchanger, "AI Has Already Decided: First-Party Data Will Define Advertising's Agentic Era" (April 2026).
7. 7. Marketing Dive, "Walmart Connect expands offsite advertising ambitions with Yahoo, Magnite" (June 2026).
8. 8. MediaPost, "Walmart Connect Expands Offsite Ad Targeting With Yahoo DSP" (May 2026).
9. 9. IAB Europe, "IAB Europe Releases Updated Commerce (Incl. Retail) Media Standards & Flexi Ad Sizes Guidelines for Retail Media Networks" (January 2026).
10. 10. IAB, "Guidelines for Incremental Measurement in Commerce Media" (November 2025).
11. 11. IAB Europe, "Measuring What Matters: Navigating the Role of Retail Media Metrics from ROAS to Incrementality" (2026).
12. 12. MarTech, "IAB finalizes retail media networks standards for in-store" (December 2024, cited as foundational standards precedent).
13. 13. Digital Shelf Institute, "What the IAB/MRC New Retail Media Guidelines Means for Brands" (April 2024, cited as foundational 2024 standards precedent).
14. 14. ppc.land, "MRC and IAB release attention measurement guidelines for advertisers" (November 2025).
15. 15. Grocery Dive, "Kroger Precision Marketing expands offsite media capabilities" (September 2025).
16. 16. Supermarket Perimeter, "Grocery retailers using Instacart's Carrot Ads see strong results" (October 2025).
17. 17. MediaPost, "Albertsons Aims To Show What In-Store Ads Actually Deliver" (January 2026).
18. 18. Digiday, "Retail media networks under scrutiny amid calls for transparency and accountability" (October 2024, quote only).
19. 19. IAB UK, "Pan-European Definitions for the Retail Media Sector" (October 2023, cited as background definitional precedent with age noted).
20. 20. Improvado, "Top 15 Retail Media Networks 2026: Rankings & Benchmarks" (July 2026). Vendor/analyst blog: improvado.io/blog/top-retail-media-networks
21. 21. Kontrol Media, "Key Metrics for Retail Media: 2026 Marketer's Guide" (June 2026). Vendor blog, single-source figure hedged in text: kontrolmedia.com/key-metrics-for-retail-media-2026-marketers-guide/
22. 22. Dataslayer, "Incrementality: Top Retail Media KPI for 2025" (November 2025, includes a secondary citation of Bain & Company). Vendor blog: dataslayer.ai/blog/incrementality-becomes-the-primary-kpi-for-retail-media-advertisers
23. 23. Happydemics, "Retail Media Analytics and Benchmarks for 2026" (December 2025). Vendor blog, single-source brand-lift figure attributed in text: happydemics.com/blog/en/retail-media-analytics-and-benchmarks-for-2026/
24. 24. Adelaide, "Adelaide Releases 2026 Outcomes Guide" (January 2026). Vendor blog, optional attention layer attributed in text: adelaidemetrics.com/blog/adelaide-releases-2026-outcomes-guide
25. 25. Retail Dive (sponsored content by Fluent, Inc.), "US Retail Media Spend Hits $69.33 Billion in 2026. Physical Stores Get Just 3.3% of It." (July 2026). Sponsored source, bias disclosed in text: eightx.co/blog/in-store-retail-media-budget-allocation-blind-spot-dtc

Note on sourcing: neutral standards bodies, analyst firms, and trade publications are hyperlinked inline at the claims they support. Platform-reported results (Walmart Connect, Kroger Precision Marketing, Instacart, Albertsons) are credited to each platform as its own reported figures and sourced through neutral trade press rather than through any platform's own domain. Vendor and analyst blogs (sources 20 to 24) are listed in plain text and were used only for directional, single-source, or optional figures, each attributed by name in the body. The IAB/MRC 2024 guidelines and the IAB UK 2023 taxonomy are cited as foundational precedent with their original publication dates noted. The 3.3% in-store figure (source 25) is sponsored content from an in-store measurement vendor and is disclosed as such; the iROAS-variance figure (source 21) is a single medium-credibility source and is hedged in text. The MRC viewability standard is referenced in general terms as long-established industry convention rather than by a specific numeric threshold.